STRUCTURAL ELEMENTS OF THE CYTOPLASM. 



189 



in my first account. These I wish now to correct, leaving 

 several further points undecided pending additional study of the 

 material as a whole. I have examined thus far the antheridia 

 of three species of Polytrichum juniperinum, piliferum and 

 commune. While some minor differences doubtless occur, for 

 practical purposes these three species may be considered as a 

 unit basis for description. 



In the androgonia and androcytes the vacuome has not yet 

 been identified, and it is uncertain whether any occurs. The 

 three other cytoplasmic components occur in clearly developed 

 form. The osmiophilic platelets (Fig. 8) are similar to those in 

 the root-tips. They tend to become smaller as the size of the 

 multiplying androgonia diminishes, and are distributed ap- 

 parently intact to the daughter cells of each division, and without 

 any obvious orientation with respect to the spindle. They 

 arrive in the androcytes as scattered bodies of the same shape 

 and staining capacity as in the androgonia. The pseudo- 

 chondriome is represented in early androgonia by scattered 

 vesicular granules (Fig. 10), exactly similar to those in root-tips. 

 In later generations of androgonia they frequently undergo 

 changes in shape, becoming rod- or thread-like. They seem to 

 arrive in this form in the androcyte, although on this point I am 

 still in some uncertainty. In cell division they show relations to 

 the spindle axis of several types which can not here be elaborated 

 upon. The plastidome occurs in the earlier androgonia in the 

 form of an extensive lamella or plate (Fig. 10), the polar plates 

 of Allen ('17). In later generations these become more compact 

 and are molded into definite rings, of which one occurs in each 

 cell. Fig. 9 shows an antheridial tip with the epithelium en- 

 closing the central mass of androgonia. The plastids in the 

 epithelium and the polar plates or rings are clearly blackened 

 by osmic acid, yielding a remarkably clear preparation. A 

 study of young antheridia in all stages of formation indicates 

 pretty clearly that these polar plates are directly descended 

 from the plastids of the cells which formed the primordia of 

 each antheridium. In the developing male cells their identity is 

 retained but their function as starch-producing plastids is 

 apparently suspended. They have a striking point in common 







